Adam Wu, Jean Michel Lauzon, Indah Andriani, and Brian R. James*

Transcription

1 Electronic Supplementary Material (ESI) for RSC Advances. This journal is The Royal Society of Chemistry 2014 Breakdown of lignins, lignin model compounds, and hydroxy-aromatics, to C 1 and C 2 chemicals via metal-free oxidation with peroxide or persulfate under mild conditions Adam Wu, Jean Michel Lauzon, Indah Andriani, and Brian R. James* Department of Chemistry, University of British Columbia, Vancouver, British Columbia, Canada, V6T 1Z1. brj@chem.ubc.ca Table of Contents General Information...S2 Experimental Section...S3 Table S1...S3 Table S2...S4 Table S3...S5 Spectroscopic Data...S6 Figure S1...S6 Figure S2...S7 Figure S3...S7 Table S4...S8 Figure S4...S9 Table S5...S10 Figure S5...S11 Table S6...S12 Figure S6...S13 Quantitative 13 C{ 1 H} Integration...S14 Reactivity at Room Temperature...S14 Decomposition of H 2 O 2...S15 References...S15 S1

2 General Information All commercially available compounds were purchased and used as received unless otherwise noted. The 31.5 wt% solution of H 2 O 2 was kept refrigerated prior to use. Dimer LMCs were synthesized according to literature procedures. 1 Lignin samples were acquired from various sources: Alkali lignin (from Aldrich) (9); pyrolytic lignin (10); a CH 2 Cl 2 -soluble fraction Kraft lignin (from Prof. J. Kadla, Faculty of Forestry, University of British Columbia) (11); indulin AT Kraft lignin (from MeadWestvaco Corp.) (12); and a lignin from Lignol Energy Corp. (13). 1 H NMR data were collected at 298 K on Bruker AVANCE 300 or 400 MHz spectrometers and were referenced to hexamethylbenzene in benzene-d 6 (δ H = 2.12) contained in a capillary inside the NMR tube. The integration error is estimated at ± 2% due to excellent peak separation and signal-to-noise ratio. Methoxy content of lignin 11 was calculated by dissolving 15 mg of the lignin and a known amount of pivalic acid in 500 μl of CDCl 3. 2 A recycle delay (d1) of 5 seconds was used to ensure complete relaxation of all proton nuclei. The mmols of OMe were calculated from the ratio of pivalic acid signals to OMe signals (δ H = ); the data were converted into a weight and subsequently a percentage (18.7%). Qualitative 13 C NMR data were collected at 298 K on the AVANCE 300 MHz spectrometer (75 MHz for 13 C) equipped with a Bruker 5 mm QNP probe at the University of British Columbia. The inverse-gated UDEFT pulse sequence 3 was used in order to increase the signal strength of quaternary carbons, specifically for the low-field formate, carbonate, and oxalate species. At least 10,000 FIDs were accumulated for each experiment resulting in run times of ~12 h. Unfortunately, quantitative data could not be collected from these runs due to instrument sensitivity and time restrictions. Quantitative 13 C NMR data were collected at 298 K on a Bruker AVANCE II 600 MHz spectrometer (150 MHz for 13 C) equipped with a Bruker 5 mm QNP cryogenically-cooled probe at Simon Fraser University (Burnaby, BC, Canada) by Dr. Andrew Lewis. Inverse-gated 1 H composite pulse decoupling (WALTZ-16) was employed with a recycle delay (d1) of 150 seconds to ensure complete relaxation of all carbon nuclei which was verified using pivalic acid as an internal standard. At least 400 FIDs were accumulated for each experiment (~16 h) in order to ensure a minimum signal-to-noise ratio of 25:1 for the formate, carbonate, and oxalate signals; this resulted in an error of about ± 2% in the integrations of the pivalic acid signals. S2

3 Experimental Section General Procedure for the Breakdown of LMCs and Lignins The appropriate amount of LMC (0.050 mmol for monomeric models; mmol for dimeric models) or lignin sample (15 mg) was added to an NMR tube with a screw-cap closure. 1 ml of 1 M KOH or 1 M K 2 CO 3 solution in D 2 O was added to the tube via syringe. A capillary standard containing hexamethylbenzene in benzene-d 6 was also added. An initial time zero 1 H NMR spectrum was recorded at 298 K at which point either H 2 O 2 (0.50 mmol, 50 μl) or K 2 S 2 O 8 (0.25 mmol, 68 mg) was added. RuCl 3 H 2 O (0.015 mmol, 3.9 mg) was sometimes added at this stage. The NMR tube was then placed in a pre-heated 60 C oil-bath for 3 h. The tube was then cooled to ambient temperature before the addition of pivalic acid ( mmol, 5-10 mg), an internal spectroscopic standard for determination of substrate consumption and product yields. The relevant 1 H NMR signals (in basic D 2 O) are presented in Table S1. Table S1 Relevant 1 H NMR signals for substrates, products, and standards. Compound δ H (ppm) Proton Multiplicity Guaiacol (1) 3.67 OCH 3 singlet Syringic Acid (2) 3.73 OCH 3 singlet Syringyl Aldehyde (3) 3.70 OCH 3 singlet Syringyl Alcohol (4) 3.66 OCH 3 singlet Veratric Acid (5) 3.77 OCH 3 singlet Vanillic Acid (6) 3.73 OCH 3 singlet LMC Dimer (7) a n/a n/a n/a Catechol Ar-H multiplet Phenol Ar-H multiplet Methanol 3.30 OCH 3 singlet Formate 8.45 HCO - 2 singlet 1.03 CH 3 singlet a The dimer is not fully soluble in the basic aqueous solutions. S3

4 Table S2 Reactions of LMCs and lignins in 1 M KOH solution (from 1 H NMR data). Substrate Oxidant Consumption (%) Methanol (mmol) Formate (mmol) Guaiacol (1) H 2 O K 2 S 2 O Syringic Acid (2) H 2 O K 2 S 2 O b b Syringyl Aldehyde (3) H 2 O K 2 S 2 O Syringyl Alcohol (4) H 2 O K 2 S 2 O Veratric Acid (5) H 2 O K 2 S 2 O Vanillic Acid (6) H 2 O K 2 S 2 O LMC Dimer (7) a H 2 O 2 n/a K 2 S 2 O 8 n/a H 2 O K 2 S 2 O H 2 O K 2 S 2 O H 2 O K 2 S 2 O c c 12 H 2 O K 2 S 2 O H 2 O K 2 S 2 O a The dimer is not fully soluble in the basic aqueous solutions and consumption cannot be accurately determined. b When RuCl 3 3H 2 O is added to form RuO 2-4, the values for methanol and formate are and mmol, respectively. c When RuCl 3 3H 2 O is added to form RuO 2-4, the values for methanol and formate are and mmol, respectively. S4

5 Table S3 Reactions of LMCs and lignins in 1 M K 2 CO 3 solution (from 1 H NMR data). Substrate Oxidant Consumption (%) Methanol (mmol) Formate (mmol) Guaiacol (1) H 2 O K 2 S 2 O Syringic Acid (2) H 2 O K 2 S 2 O Syringyl Aldehyde (3) H 2 O K 2 S 2 O Syringyl Alcohol (4) H 2 O K 2 S 2 O Veratric Acid (5) H 2 O K 2 S 2 O Vanillic Acid (6) H 2 O K 2 S 2 O b H 2 O 2 n/a K 2 S 2 O 8 n/a b H 2 O 2 n/a K 2 S 2 O 8 n/a b H 2 O 2 n/a K 2 S 2 O 8 n/a b H 2 O 2 n/a K 2 S 2 O 8 n/a b H 2 O 2 n/a K 2 S 2 O 8 n/a a The dimer is not fully soluble in the basic aqueous solutions and consumption cannot be accurately determined. b Lignins only partially dissolve in the 1 M K 2 CO 3 solutions and consumption cannot be accurately determined. S5

6 Spectroscopic Data Chemical Shift (ppm) Figure S1 1 H NMR spectra for a typical reaction mixture. In this example, syringic acid (2) is reacted with H 2 O 2 in 1 M KOH according to the General Procedure. The bottom spectrum (black) is the initial time zero spectrum; the top spectrum (red) was taken after 3 h reaction at 60 C (after addition of pivalic acid). See Table S1 for signal assignment. In addition to the residual substrate signals and those of methanol and formate, two trace, unassigned signals (δ H = 3.61 and 4.99) are seen for the syringyl derived substrates (2 4). 1 H 13 C NMR spectroscopy (HSQC, Fig. S2) of a completed reaction for syringic acid (2) reveals a relationship between the δ H = 3.61 and δ C = 55.7 signals. Integrations from quantitative 13 C NMR (vide infra) and the corresponding 1 H NMR data suggest these signals are representative of a CH 3 group. The HSQC spectrum also shows the δ H = 4.99 and δ C = 95.0 signals are related; however, the integrations are uninformative about the number of protons attached to the carbon. A COSY 1 H 1 H NMR experiment (Fig. S3) confirms that the δ H = 3.61 and 4.99 signals are related, suggesting a molecule with at least a two carbon chain, but this side-product has not yet been identified. S6

7 4.99 x x F1 Chemical Shift (ppm) F2 Chemical Shift (ppm) Figure S2 HSQC spectrum for syringic acid (2) reacted with H 2 O 2 in 1 M KOH according to the General Procedure. The correlations shown from right to left are as follows: δ H = 3.30 and δ C = 48.7 (methanol); δ H = 3.61 and δ C = 55.7 (unidentified); δ H = 4.99 and δ C = 95.0 (unidentified); δ H = 8.45 and δ C = (formate) x x F1 Chemical Shift (ppm) F2 Chemical Shift (ppm) Figure S3 COSY spectrum for syringic acid (2) reacted with H 2 O 2 in 1 M KOH according to the General Procedure. The relevant off-diagonal signals are highlighted for clarity. S7

8 Table S4 Quantitative 13 C{ 1 H} data for the reaction of syringic acid (2) with H 2 O 2 in 1 M KOH according to the General Procedure. Compound δ C (ppm) Amount (mmol) % Total Carbon Present (3 carbons) Internal Standard Internal Standard Methanol Unidentified Unidentified Unidentified Carbonate Formate Unidentified Oxalate (2 carbons) Unidentified Unidentified Unidentified Internal Standard Total Identified Carbon (MeOH + HCO CO C 2 O 4 2- ) = mmol Total Unidentified Carbon (see italics) = mmol Total Carbon Present (identified + unidentified) = mmol Expected Carbon = mmol Total Identified / Total Carbon Present = 76.3% Total Unidentified / Total Carbon Present = 23.7% Total Carbon / Expected Carbon = 73.7% Integration of the 1 H NMR spectrum (600 MHz) of the same sample gives the following: mmol MeOH (1.4% change from 13 C data) mmol HCO 2 - (6.8% change from 13 C data) S8

9 Methanol Chemical Shift (ppm) Carbonate Oxalate Formate Chemical Shift (ppm) Figure S4 13 C{ 1 H} spectrum corresponding to the data in Table S4. S9

10 Table S5 Quantitative 13 C{ 1 H} data for the reaction of syringyl aldehyde (3) with H 2 O 2 in 1 M KOH according to the General Procedure. Compound δ C (ppm) Amount (mmol) % Total Carbon Present (3 carbons) Internal Standard Internal Standard Methanol Unidentified Unidentified Unidentified Carbonate Formate Unidentified Oxalate (2 carbons) Unidentified Unidentified Unidentified Internal Standard Total Identified Carbon (MeOH + HCO CO C 2 O 4 2- ) = mmol Total Unidentified Carbon (see italics) = mmol Total Carbon Present (identified + unidentified) = mmol Expected Carbon = mmol Total Identified / Total Carbon Present = 72.5% Total Unidentified / Total Carbon Present = 27.5% Total Carbon / Expected Carbon = 87.3% Integration of the 1 H NMR spectrum (600 MHz) of the same sample gives the following: mmol MeOH (0.0% change from 13 C data) mmol HCO 2 - (2.5% change from 13 C data) S10

11 Methanol Chemical Shift (ppm) Formate Oxalate Carbonate Chemical Shift (ppm) Figure S5 13 C{ 1 H} spectrum corresponding to the data in Table S5. S11

12 Table S6 Quantitative 13 C{ 1 H} data for the reaction of a CH 2 Cl 2 -soluble fraction Kraft lignin (11) with H 2 O 2 in 1 M KOH according to the General Procedure. Compound δ C (ppm) Amount (mmol) % Total Carbon Present (3 carbons) Internal Standard Internal Standard Methanol Unidentified Unidentified Unidentified Carbonate Formate Unidentified Oxalate (2 carbons) Unidentified Unidentified Unidentified Internal Standard Total Identified Carbon (MeOH + HCO CO C 2 O 4 2- ) = mmol Total Unidentified Carbon (see italics) = mmol Total Carbon Present (identified + unidentified) = mmol Total Identified / Total Carbon Present = 86.4% Total Unidentified / Total Carbon Present = 13.6% Integration of the 1 H NMR spectrum (600 MHz) of the same sample gives the following: mmol MeOH (3.7% change from 13 C data) mmol HCO 2 - (26.9% change from 13 C data) S12

13 Methanol Chemical Shift (ppm) Carbonate Formate Oxalate Chemical Shift (ppm) Figure S6 13 C{ 1 H} spectrum corresponding to the data in Table S6. S13

14 Quantitative 13 C{ 1 H} Integration The range of total carbon detected by 13 C{ 1 H} referred to in the main text arises from differences in the integration method of the signals. The spectra were processed using two different software suites (ACD Labs and Topspin 2.1) and were integrated by hand and by region definition. Though the values for the larger signals (pivalic acid, methanol, formate, carbonate, oxalate) showed only small variations, the 7 unidentified signals experienced larger deviations due to their size relative to the baseline. The signal-to-noise ratio of 25:1 gives excellent data for the internal standard (± 2%) but is not precise enough to give exact data on the unidentified signals that are just 3 times the baseline height. The values reported in Tables S4, S5, and S6 are integrated by hand using ACD Labs and represent the lower threshold of the reported ranges. Reactivity at Room Temperature Of note, no substrates signals are present in the quantitative 13 C{ 1 H} NMR spectra reported herein. This is a result of the r.t. reactivity described in the text. After the 3 h reaction time with H 2 O 2 at 60 C, a 1 H NMR spectrum was taken as outlined in the General Procedure and the substrate consumptions were calculated (94 and 100% for 2 and 3 respectively); consumption of 11 was not calculated due to poor shimming that resulted in overlap between the remaining aromatic signals and the C 6 D 6 signal). In the quantitative 13 C{ 1 H} NMR experiments performed at Simon Fraser University, the samples were allowed to sit at r.t. overnight or even longer depending on availability of the spectrometer. However, before beginning the 13 C{ 1 H} NMR analysis a 1 H NMR spectrum was taken and these showed no substrate signals, suggesting complete consumption before the quantitative experiments began and thus explaining the lack of corresponding signals in the 13 C{ 1 H} NMR spectra. Although our studies involving r.t. reactivity of LMCs and lignins are preliminary, the product ratios of MeOH to formate (as determined by 1 H NMR) vary little between reactions at 60 C and those performed at r.t. Therefore, it is supposed that the values gleaned from the quantitative 13 C{ 1 H} NMR analysis are accurate as r.t. reactivity accounts for < 10% of the overall substrate consumption. S14

15 Decomposition of H 2 O 2 In an effort to probe the stoichiometry of reactions, the possibility of determining the amount of H 2 O 2 that remained after the 3 h was explored. However, titration of H 2 O 2 with standard KMnO 4 or iodine solutions requires acidic media whereas the degradation reactions require a basic environment. To this end, 50 μl of 31.5 wt. % H 2 O 2 solution was dissolved in 1 ml of 1 M KOH solution and immediately neutralized with 1 M HCl before undergoing titration; such samples had the same [H 2 O 2 ] (within error) as those taken directly from the commercial source. Conversely, H 2 O 2 samples that were left in the basic solution for 1 h or shaken vigorously for 5 min had, on average, only ~73% of the value found in untreated samples. Peroxide samples heated at 60 C for 2 h in the basic solution reveal a dramatic decrease in [H 2 O 2 ], with only ~18% remaining; the findings are consistent with the known H 2 O 2 decomposition to O 2 and water at higher ph and/or temperature. 4 The variable decomposition of the H 2 O 2 is considered to be the main source of error (± 15%) in consumption values for the degradation of the LMCs. References 1 A. Wu, B. O. Patrick, E. Chung and B. R. James, Dalton Trans., 2012, 41, N.-E. El Mansouri, Q. Yuan and F. Huang, BioRes., 2011, 6, M. Piotto, M. Bourdonneau, K. Elbayed, J.-M. Wieruszeski and G. Lippens, Magn. Reson. Chem., 2006, 44, F. R. Duke and T. W. Haas, The Journal of Physical Chemistry, 1961, 65, S15